The purpose of the proposed continuation phase of this investigation is to capitalize on what has been learned about the Personality Disorder Examination (PDE) during the first year of support, make suitable adjustments for the final revision of DSM- III and the introduction of ICD-10, and enlarge the scope of the original proposal. The revised PDE will not only reflect the changes in the two nomenclatures, but it will also incorporate whatever modifications seem appropriate based on our experience thus far, including data analysis of more than 100 cases from the current clinical trial. The specific aims of the research are: (1) To revise the PDE to reflect the disorders and criteria in the final version DSM-III-R as well as improvements suggested by data collected in a sample of more than 100 patients. (2) To make a similar revision in the international version of the PDE to accommodate the new criterion-based ICD-10 nomenclature, and to participate as an American site in the initial WHO/ADAMHA pilot study of the PDE in Europe, Africa, Asia, and North America. (3) To determine the interrater agreement on the revised PDE for the categorical diagnoses, dimensional scores, and criteria. (4) To determine the test-retest reliability of the revised PDE over significant time intervals for the categorical diagnoses, dimensional scores, and criteria. (5) To develop a glossary and detailed, comprehensive scoring manual for the PDE. (6) To identify the subject and interviewer variables affecting the test-retest reliability of the PDE. (7) To compare the PDE and the Personality Disorder Inventory (PDI), a self-administered test of personality disorders. (8) To compare the patient and informant versions of the PDE and PDI, and to develop procedures for integrating the two sources of information and thus enhancing the validity of the PDE. (9) To determine the reliability, sensitivity, specificity, and positive and negative predictive values of the DSM-III-R criteria and individual items on the PDE for each of the DSM-III-R personality disorders, including an analysis of the influence of age and sex. (10) To apply cluster analytic techniques to identify the syndromal sampled by Axis II, and to use causal modeling methods to determine whether there is mathematical support for the DSM-III-R taxonomy.